tjuptz 发表于 2021-8-11 22:32
从官方论坛来看,常加速md这个功能应该是被砍了,因为有bug年久失修
I did an in depth analysis of this because I got told by several users that this was working as designed.
@mark.j.abraham is of course correct that the call to update_ekindata has been gone for ages, but the code takes the acceleration directly from the inputrec.opts.acc rvec and adds it to the actual velocity at src/gromacs/mdlib/update.cpp:610.
This means that what is supposed to be computed for the acceleration is skipped (the grpstat[].u field is always zero and never changed), and the velocity is simply scaled by whatever the user set in the accelerate mdp option.
lyj714 发表于 2021-8-11 22:34
不是吧,啊sr。我测试过各种版本gmx(2016,2018,2019,2020,2021)都能用,你是不是命令写错了。Bug是有,只 ...
tjuptz 发表于 2021-8-12 22:28
刘老师,可否帮忙看一下,十分感激!刚刚发错楼了,忘记用回复了
lyj714 发表于 2021-8-12 22:41
mdp中有奇怪字符,你的accelerate = 0.1 0.0 0.0 -0.1 0.0 0.0的数值不是空格隔开的,自己重新打一下就 ...
萌新杨同学 发表于 2023-4-30 21:06
老师您好,首先感谢您分享了非平衡MD的mdp文件,但是我有一个疑虑:您对两个组分别施加了恒定的加速度,那 ...
tjuptz 发表于 2023-5-2 08:57
不能,所以基本没人用gromacs 做纳米流动
萌新杨同学 发表于 2023-5-2 10:56
感谢前辈的回复~
还想再向前辈请教一下可否在控温时只对不施加外力的组进行控温而对施加外力的组不控温 ...
tjuptz 发表于 2023-5-2 15:36
你可以看一下gmx手册里是否提到施加加速度后的控温器是不是已经排除了加速度组,如果不是可以自己尝试比 ...
萌新杨同学 发表于 2023-5-2 16:43
好的,感谢您的回复,我再研究研究-----------------------------
gmx手册中指出被加速的组也会对系统 ...
欢迎光临 计算化学公社 (http://bbs.keinsci.com/) | Powered by Discuz! X3.3 |